Cannot install: ERROR: One or more destination directories are not accessible !!!

[ritog]

I see this error:

ERROR: One or more destination directories are not accessible !!!

And I am unable to install J kernel on my machine.

---

Here is the full error message:

(j) ~/Applications/jkernel-master> python setup.py install

Check pre-requisites ...
Anaconda/Miniconda root directory ... : /home/ritog/.local OK.
Site Packages directory ............. : /home/ritog/.local/lib/python3.10/site-packages OK.
Jupyter Notebook syntax directory ... : /home/ritog/.local/lib/python3.10/site-packages/notebook/static/components/codemirror/mode MISSING or NOT writeable.
Jupyter Kernels directory ........... : /home/ritog/.local/share/jupyter/kernels MISSING or NOT writeable.
ERROR: One or more destination directories are not accessible !!!

How can I solve this problem?

---

Note: I have a proper J installation that works through the ijconsole, jconsole, and jsx (on browser).

[martin-saurer]

May I ask you a few questions?:

• Are you on a Mac?
• Which "python" are you executing on the command line? The one from MacOS or the one from Anaconda?

I recommend installing Anaconda into your home directory: /home/ritog/anaconda3 Then execute the command line: /home/ritog/anaconda3/bin/python setup.py install setup.py builds all the necessary paths relative to the "python" executable.

Please let me know, if that helped.

[ritog]

Hi @martin-saurer , I am not on Mac, but on Linux (Pop OS 22.04 LTS, 64 bit).

I am using miniconda, and it is indeed in /home/ritog/miniconda3.

I have created a conda environment, and activated it and installed jupyter in it, and I am running python from within that environment.

[martin-saurer]

Hi @ghosh-r , I have installed Miniconda, created an environment, installed jkernel, and everything worked right out-of-the-box. I'm on Linux Mint 20.3.

$ conda env list
# conda environments:
#
                         /home/martin/Anaconda3
                         /home/martin/Anaconda3_NLP
                         /home/martin/Anaconda3_oTree
base                     /home/martin/Miniconda3
jj                    *  /home/martin/Miniconda3/envs/jj

$ which python
/home/martin/Miniconda3/envs/jj/bin/python

$ cd /home/martin/Program/jkernel
$ python setup.py check
Check pre-requisites ...
Anaconda/Miniconda root directory ... : /home/martin/Miniconda3/envs/jj OK.
Site Packages directory ............. : /home/martin/Miniconda3/envs/jj/lib/python3.10/site-packages OK.
Jupyter Notebook syntax directory ... : /home/martin/Miniconda3/envs/jj/lib/python3.10/site-packages/notebook/static/components/codemirror/mode OK.
Jupyter Kernels directory ........... : /home/martin/Miniconda3/envs/jj/share/jupyter/kernels OK.
ALL destination directories are OK.
Done.

Checking current installation ...
/home/martin/Miniconda3/envs/jj/lib/python3.10/site-packages/jkernel => OK
/home/martin/Miniconda3/envs/jj/lib/python3.10/site-packages/jkernel/__init__.py => OK
/home/martin/Miniconda3/envs/jj/lib/python3.10/site-packages/jkernel/jinter.py => OK
/home/martin/Miniconda3/envs/jj/lib/python3.10/site-packages/jkernel/jkernel.py => OK
/home/martin/Miniconda3/envs/jj/share/jupyter/kernels/jkernel => OK
/home/martin/Miniconda3/envs/jj/share/jupyter/kernels/jkernel/kernel.json => OK
/home/martin/Miniconda3/envs/jj/share/jupyter/kernels/jkernel/logo-32x32.png => OK
/home/martin/Miniconda3/envs/jj/share/jupyter/kernels/jkernel/logo-64x64.png => OK
/home/martin/Miniconda3/envs/jj/lib/python3.10/site-packages/notebook/static/components/codemirror/mode/j => OK
/home/martin/Miniconda3/envs/jj/lib/python3.10/site-packages/notebook/static/components/codemirror/mode/j/j.js => OK
Environment variable: J_INSTALLATION_FOLDER => OK
/home/martin/J903 => OK
Environment variable: J_BIN_FOLDER => OK
/home/martin/J903/bin => OK
ALL checks passed.
Done.

There must be something different in your installation, but I cannot see what it is. It looks strange to me, that in your environment the directory .../notebook/static/components/codemirror/mode seems to be missing.

Does Jupyter Notebook (Python Kernel) work in your environment?

[ritog]

Does Jupyter Notebook (Python Kernel) work in your environment?

Yes, @martin-saurer it does. Jupyter Notebook works fine, and also I can run it with a Python3 kernel.

It might be of interest to you that after I failed to install this program, I went ahead with the J download that comes bunched up with Jupyter.

That seems to work out fine.

I downloaded the file from J software website's Jupyter download page. And followed the instruction provided on the Jupyter Guide page.

[martin-saurer]

Fixed in setup.py